system is 2-node-cluster based on OES2 Linux (SP0). During testing the
cluster failover functions I found out that a cluster node will continue
living when the local file system is unavailable (e.g. all local hard
disks have failed). This means that cluster services are still running
on the machine, but the ressources in fact become unusable (clusterd nss
volume etc).

The node does *not* die and therefor the cluster ressources do *not*
failover to a surviving node!

How can I solve this problem? My first idea is to have some kind of
service that checks the local file system and resets the server if it is
unavailable. Resetting the server would cause the cluster resources to

Is there such a service available or which other method exist to solve
the problem?

Thanks in advance,